PHYLOGENY OF THE GENUS RAPHIONACME
(APOCYNACEAE)
by
MAGDIL PIENAAR
Submitted in fulfilment of the requirements for the degree
MAGISTER SCIENTIAE
in the Faculty of Natural and Agricultural Sciences,
Department of Plant Sciences (Botany)
at the
University of the Free State
Bloemfontein
July 2013
Supervisor: Dr B. Visser
Co-supervisor: Dr A.M. Venter
Co-supervisor: Dr M. Jackson
i
ACKNOWLEDGEMENTS
I would like to show my sincere gratitude and appreciation to the following persons:
My family, especially my dad and mom, Stoffel and Marcia Pienaar for their never-ending support and for believing that I can achieve anything! Also thanks to my brother and sister, André and Chalandi Pienaar.
Prof Johan Venter and Dr Andor Venter, for their help with fieldwork and their absolute exceptional botanical knowledge. Without your help I would not have been able to complete this study.
My supervisor and co-supervisors, Dr Botma Visser, Dr Andor Venter and Dr Mariette Jackson for all your patience during the duration of this study.
My friends for all your support and happy faces. Especially Dr Lize Joubert for your help during the fieldwork and your support and input.
Last but not least I would like to thank all the herbaria from which I was able to borrow herbarium material. Without the specimens this study would not have been a success.
ii
LIST OF CONTENTS
Acknowledgements i
List of abbreviations vi
List of tables and figures ix
List of Raphionacme species xi
Chapter 1: Introduction
1.1. General introduction 1
1.2. Aim 2
Chapter 2: Literature review
2.1. Biogeography of Africa 4
2.2. The family Apocynaceae 17
2.3. The subfamily Periplocoideae 19
2.3.1. Circumscription of the Periplocoideae 19
2.3.2. Distribution and habitat of the Periplocoideae 21
2.4. The genus Raphionacme 22
2.4.1. Historical notes on Raphionacme 22
2.4.2. Sectional classification of Raphionacme 23
iii
3.4.4. Morphology of Raphionacme 25
2.5. Molecular analysis 27
2.5.1. Ribosomal DNA (rDNA) 27
2.5.2. Characterising the Internal Transcribed Spacer (ITS) 27
2.5.3. Complications with ITS 30
2.5.4. The value of the ITS region 31
2.6. Morphological analysis 33
2.6.1. The value and application of morphological data 33 2.6.2. Problems with ad restrictions of morphological data 35
2.7. Phylogenetics 36
Chapter 3: Material and Methods
3.1. Materials 38
3.1.1. Specimens 38
3.1.2. Morphological characters and character states 38
3.1.3. Morphological terminology 38
3.1.4. Biogeography 44
3.2. Methods 44
3.2.1. DNA extraction from dried leaf tissue 44 3.2.2. Sequence analysis of plant specimens 45 3.2.2.1. Amplification of the ITS region 45
3.2.2.2. DNA sequencing of ITS 48
iv
3.2.3. Phylogenetic analyses 49 3.2.3.1. Data 49 3.2.3.2. Parsimony 49 3.2.3.3. Maximum likelihood 51 3.2.3.4. Bayesian analysis 51 3.2.4. Biogeography 51 Chapter 4: Results 4.1. Biogeography 52 4.1.1. Distribution of Raphionacme 524.1.2. Raphionacme species richness 52
4.1.3. Endemism in Raphionacme 55
4.2. Cladistical analyses 55
4.2.1. Morphological data 55
4.2.2. Molecular data 58
4.2.3. Combined data 60
Chapter 5: Discussion and Conclusion
5.1. Incongruence between the morphological and nuclear datasets 62 5.2. Phylogeny, morphology and biogeography of Raphionacme 62 5.2.1. Phylogenetic relationships and biogeographical affinities 62 in Raphionacme
v
5.3.1. Forms of Raphionacme hirsuta 77
5.3.2. Raphionacme splendens – unite or divide 77 5.3.3. Sectional classification of Raphionacme 82 5.3.4. Position of Schlechterella abyssinica 82 5.4. Distribution of endemism and species richness 84 5.4.1. Distribution of Raphionacme species within 85 Regions of Endemism
5.4.2. Species richness 90
5.5. Possible origin and distribution of Raphionacme 91
5.6. Conclusion 94
References 97
Appendix 116
Summary 121
vi
LIST OF ABBREVIATIONS
A
A Adenine
AIC Akaike Information Criterion
B
BI Bayesian Analysis / Bayesian Inference
bp Base Pairs
BRAHMS Botanic Research And Herbarium Management System
BS Bootstrap
C
C Cytosine
CI Consistency Index
CTAB Cetyltrimethylammonium bromide
D
DNA Deoxyribonucleic acid
E
EDTA Ethylenediaminetetraacetic acid
G
vii
HHI Homoplasy Index
I
ICBN International Code of Botanical Nomenclature
ILD Incongruence Length Difference
ITS Internal Transcribed Spacer Region
M
MCMC Markov Chain Monte Carlo
ML Maximum Likelihood
MP Maximum Parsimony
mRNA Messenger RNA
N
nrDNA Nuclear Ribosomal DNA
O
OG Outgroup
ORD Orbitally Forced Range Dynamics
P
PCR Polymerase Chain Reaction
viii
RrDNA Ribosomal DNA
RI Retention Index RNase Ribonuclease A T T Thymine TAE Tris-acetate-EDTA TBR Tree-Bisection-Reconnection
Tris-HCl Tris(hydoxymethyl)aminomethane – hydrochloride
U
UV Ultraviolet
Y
ix
LIST OF TABLES AND FIGURES
Table 2.1. Classification of the main phytochoria of Africa (White, 1983). 10
Table 2.2. Raphionacme species within Venter and Verhoeven’s (1988) four sections.
24
Table 3.1. List of morphological characters with character states as used in the data-matrix.
39
Table 3.2. Nucleotide sequences of ITS primers used in this study. 47
Table 3.3. Raphionacme and outgroup species with their respective accession numbers for which ITS sequences were obtained from GenBank.
50
Table 4.1. Raphionacme species endemic to White’s (1983) phytochoria. 56
Fig. 2.1. White’s (1983) vegetation map of Africa indicating the main phytochoria.
11
Fig. 2.2. Schematic representation of the universal structure of the rDNA region in plants as adopted from Poczai and Hyvönen (2010).
28
Fig. 3.1. Schematic representation of a generalized Raphionacme flower. 42
Fig. 3.2. Schematic representation of ITS regions 1 and 2 as adopted from Blattner (1999).
46
Fig. 4.1. Known distribution of Raphionacme species in Africa. 53
x
Fig. 4.3. Morphological data cladogram of Raphionacme species. 57
Fig. 4.4. ITS cladogram of Raphionacme species. 59
Fig. 4.5. Cladogram based on combined data of ITS and morphology of Raphionacme species.
61
Fig. 5.1. Distribution of Raphionacme species within clade A. 64
Fig. 5.2. Distribution of Raphionacme species within sub-clade A1. 66
Fig. 5.3. Distribution of Raphionacme species within sub-clade A2. 67
Fig. 5.4. Distribution of Raphionacme species within sub-clade A3. 69
Fig. 5.5. Distribution of Raphionacme species within sub-clade A4. 70
Fig. 5.6. Distribution of Raphionacme species within clade B. 71
Fig. 5.7. Distribution of Raphionacme species within clade C. 73
Fig. 5.8. Distribution of Raphionacme species within sub-clade C1. 75
Fig. 5.9. Distribution of Raphionacme species within clade D. 76
Fig. 5.10. Voucher specimens of the two Raphionacme hirsuta forms. 78
Fig. 5.11. ITS sequence alignment of the two Raphionacme hirsuta forms. 79
Fig. 5.12. Specimens of (i) Raphionacme splendens subsp. bingeri and (ii) R. splendens subsp. splendens.
80
Fig. 5.13. Sections of Raphionacme plotted onto the combined data cladogram.
xi
LIST OF RAPHIONACME SPECIES NAMES
Raphionacme arabica A.G.Mill. & Biagi
Raphionacme angolensis (Baill.) N.E.Br. = Raphionacme kubangensis
Raphionacme bingeri (A.Chev.) Lebrun & Stork = Raphionacme splendens subsp. bingeri
Raphionacme borenensis Venter & M.G.Gilbert Raphionacme brownii Scott-Elliott
Raphionacme caerulea E.A.Bruce
Raphionacme chimanimaniana Venter & R.L.Verh. Raphionacme dyeri Retief & Venter
Raphionacme elsana Venter & R.L.Verh.
Raphionacme excisa Schltr. = Raphionacme splendens subsp. splendens Raphionacme flanaganii Schltr.
Raphionacme galpinii Schltr. Raphionacme globosa K.Schum. Raphionacme grandiflora N.E.Br.
Raphionacme haeneliae Venter & R.L.Verh. Raphionacme hirsuta (E.Mey.) R.A.Dyer Raphionacme inconspicua H.Huber Raphionacme keayii Bullock
Raphionacme kubangensis S.Moore Raphionacme lanceolata Schinz Raphionacme linearis K.Schum. Raphionacme longifolia N.E.Br. Raphionacme longituba E.A.Bruce Raphionacme lucens Venter & R.L.Verh.
xii
Raphionacme madiensis S.MooreRaphionacme michelii De Wild.
Raphionacme monteiroae (Oliv.) N.E.Br. = Chlorocyathus monteiroae Oliv. Raphionacme moyalica Venter & R.L.Verh.
Raphionacme namibiana Venter & R.L.Verh. Raphionacme palustris Venter & R.L.Verh. Raphionacme procumbens Schltr.
Raphionacme pulchella Venter & R.L.Verh. Raphionacme splendens Schltr.
Raphionacme splendens Schltr. subsp. splendens
Raphionacme splendens subsp. bingeri (A.Chev.) Venter Raphionacme sylvicola Venter & R.L.Verh.
Raphionacme utilis N.E.Br. & Stapf Raphionacme velutina Schltr. Raphionacme vignei E.A.Bruce Raphionacme villicorona Venter
Raphionacme welwitschii Schltr. & Rendle Raphionacme zeyheri Harv.
CHAPTER 1
INTRODUCTION
1
1.1. General introductionThe Gentianales, originally described by Jussieu (1789), consists of the families Apocynaceae, Gelsemiaceae, Gentianaceae, Loganiaceae and Rubiaceae. Several common vegetative, floral and phytochemical characteristics are shared by these families (APG, 2009). The vegetative forms range from small alpine herbs to large, woody, rainforest trees with opposite, entire leaves, often with stipules and colleters. Jussieu (1789) was also the author of the family name Apocineae (now known as the Apocynaceae) as part of his new order. Brown (1810) divided the Apocynaceae and named his new family the Asclepiadeae (currently known as Asclepiadaceae). However, the Apocynaceae and Asclepiadaceae share more similarities with each other than with the rest of the Gentianales families. In a number of characters there are gradations of characteristics from the Apocynaceae to the Asclepiadaceae (Endress, 2001). This made the delimitation of the Apocynaceae/Asclepiadaceae problematic and the status of these families has been the subject of on-going controversy. The most compelling evidence for uniting the Apocynaceae and Asclepiadaceae was obtained from detailed and extensive morphological studies as well as the rapidly growing body of molecular information (Judd et al., 1994; Civeyrel, 1996; Endress et al., 1996; Sennblad and Bremer, 2000; Endress, 2001; Endress and Stevens, 2001). At present the combined Apocynaceae, second largest family in the Gentianales (APG, 2009), consists of approximately 5100 species distributed among 357 genera (Nazar et al., 2013). The Apocynaceae is one of four families occurring in South Africa that are characterised by the presence of latex, the others being Euphorbiaceae, Moraceae and Sapotaceae (Leistner, 2005) and one of over 40 plant families worldwide containing latex (Hunter, 1994). The primary function of latex is protection against browsing (Hunter, 1994).
The subfamily name Periplocoideae was established when Brown (1810) divided his Asclepiadeae and named one of the groups Periploceae (Endress, 2001; 2004; Endress and Bruyns, 2000). Distribution of this subfamily is entirely Old World, in Madagascar, Asia, Australia, but is most diverse in Africa (Venter, 2009). It has been estimated that the Periplocoideae diversified in the middle Eocene (53 – 33.7 million years ago).
2
This estimated age of biogeographical diversification of Asclepiadoideae was basedon the plastid sequences trnL intron and trnL-F intergenic spacer (trnL-F). This study included one Periplocoideae species (Rapini et al., 2007).
Raphionacme Harv., comprising 36 species, is the largest genus in the Periplocoideae and is distributed throughout Africa with one species occurring in Arabia (Dyer, 1975; Venter, 2009). Economically Raphionacme does not make any significant contribution, the exception being R. utilis from which rubber was extracted commercially in the early 1900’s. However, members of this genus have numerous ethnobotanical applications. The tuber is mostly used as a source of water, medicine and food, but if applied incorrectly, this could be poisonous (Venter, 2009). Some examples are the tubers of R. hirsuta of southern Africa, which have been used to treat cancer (Graham et al., 2000) and a decoction of R. splendens leaves which are used to treat conjunctivitis (Venter, 2009). Raphionacme species of which tubers are eaten raw by locals include R. arabica, R. brownii, R. longifolia, R. madiensis, R. splendens, R. velutina and R. vignei (Miller and Biagi, 1988; Venter, 2009).
1.2. Aim
The African genera of the Periplocoideae were revised by Venter and Verhoeven (2001) and have been the subject of a number of publications. These publications included the revision of Ectadium E.Mey. (Venter et al., 1990b), Tacazzea Decne. (Venter et al., 1990a), Stomatostemma N.E.Br. (Venter and Verhoeven, 1993), Buckollia Venter & R.L.Verh. (Venter and Verhoeven, 1994a), Maclaudia Venter & R.L.Verh. (Venter and Verhoeven, 1994b), Periploca L. (Venter, 1997), Schlechterella K.Schum. (Venter and Verhoeven, 1998), Chlorocyathus Oliv. (Venter, 2008), Baseonema Schltr. & Rendle (Venter and Verhoeven, 2009), Batesanthus N.E.Br. (Venter and Verhoeven, 2009) and Raphionacme (Venter, 2009).
Regarding Raphionacme itself, no publication on the phylogeny, which includes all the species of this genus, has appeared. A small number of Raphionacme species were incorporated in some phylogenetic studies on the Apocynaceae and Periplocoideae.
3
Analyses published by Lahaye et al. (2007) contained only one Raphionacmespecies, Ionta and Judd (2007) included three species, Meve and Liede (2004) seven species and Ionta (2009) sixteen species.
Because of the complexity of the flower, especially the androecium-gynoecium complex or gynostegium, morphological studies and deductions based on a single or a few selected characteristics led to misconceptions. Unnatural grouping (Venter and Verhoeven, 1988) and circumscription of genera and species thus occurred, as was found by García et al. (2009) in their study of the Malva L. alliance.
The only way to resolve these problems of relationships between species and related genera and to divide the genus into natural groups is to add molecular data to the cladistical analyses. The phylogenetic study of the genus Raphionacme presented here complements the revision of the genus (Venter, 2009) and attempts to determine the relationships between species, evolutionary trends, distribution patterns and habitat preferences and infer a possible classification of the species. The knowledge of biodiversity is necessary in order to make informed decisions to protect habitat and biodiversity against human exploitation and destruction. Sustainable use of natural resources also depends on the knowledge of local biodiversity. Taxonomy thus provides a basic understanding of the components of biodiversity which is necessary for effective decision making about conservation and sustainable use.
Theodosius Dobzhansky stated that, “Nothing in biology makes sense except in the light of evolution”. Thus, in arrangement with explicit methods for phylogenetic analysis, molecular data have reformed concepts of relationships and limitations at all levels of the taxonomic hierarchy.
CHAPTER 2
LITERATURE REVIEW
4
2.1. Biogeography of AfricaAfrica is the second largest continent on earth (White, 1983), with many features in plant life that combine to make the continent enthralling and puzzling. The current landscape of Africa is the result of tectonic movement, since the far distant breakup of Gondwana. Another influence is the incidence of on-going episodes of rifting and volcanism since the establishment of Africa as a distinct continent. These tectonic and thermal mechanisms clarify certain specific characteristics of present-day Africa (Summerfield, 1996).
Since the early Permian, 295 to 250 million years ago, Africa has experienced no less than seven major rifting episodes (Lambiase, 1989). The first of the rifting episodes preceded the breakup of Gondwana and was the most noteworthy, relating to the Karoo Rift and Basin development. Sedimentation in the Karoo Basin of southern Africa was nearly continuous from 280 to 100 million years ago, while the uplift of the Cape Fold Belt to the south was coexisting with the early stages of the basin subsidence (Summerfield, 1996).
Rifts and basins of the early Permian originated over a wide area of Africa, ranging from the Benue Trough in the west, to Sudan in the east and the Sirte Basin in Libya to the north. Due to the permanency of these rift systems, it has been collectively termed the Central African Rift System (Fairhead, 1986). A linear feature is shaped by the rift structures, but individual basins are located at highly variable angles to the overall inclination of the rift system. The early Miocene (approximately 23.5 million years ago) involved the start of the present rifting episode (Summerfield, 1996), while the East African Rift System continues to develop to the southwest towards the Kalahari Craton as shown by seismic data (Fairhead and Reeves, 1977).
Associated with the East African Rift System are the most prominent upwarps, which itself, appears to signify the site of developing continental rifting. One such example is the series of broad upwarps rising to altitudes of several hundred metres or more above the adjacent terrain that occur over much of the continent. These are mostly evident as long-recognized marginal upwarps which run parallel to the coastline in numerous areas and which are bound on their seaward side by a sharp topographic gap in the form of a major escarpment, or series of escarpments (Summerfield, 1996).
5
Elevation distribution across the African continent is of considerable importance. Aheight-frequency distribution assessment for the whole African continent indicates a concentration of elevation between 400 and 600 m, with a secondary peak in distribution between 800 and 1000 m. The consequence of such height-area distribution lies in the way they reflect, at an essential level, the tectonic and landscape evolution of the continent (Summerfield, 1996).
First-order landscape morphological distribution within the African continent is between the broadly elevated south and east, and the lower-lying north and west (King, 1967). Africa can be divided into high and low parts when a line is drawn across the map of Africa, from Angola to western Ethiopia. Low Africa, in the north-west, consists of sedimentary basins and upland plains mostly between 150 and 600 m above sea level, comprising the Sahara and the catchments of the lower Nile, Senegal, Niger, Chad and Zaire rivers. Land rising above 1000 m is restricted mainly to the Atlas Mountains in the Maghreb (northern part of Morocco, Algeria and Tunisia), the Sahara massifs, Ahaggar and Tibetsi, Jebel Marra in the Sudan Republic, the headwaters of the Niger, the Jos Plateau in Nigeria and the Cameroon highlands. To the south and east, almost all of High Africa rises above 1000 m, with the exception of Somalia, the broad lowlands of Mozambique as well as moderately narrow coastal plains and valley strips elsewhere. Even the Kalahari basin is about 1000 m above the sea and in east Africa the surface of Lake Victoria stands 1130 m above sea-level (White, 1983).
Africa’s tectonic history did not result in the formation of great cordillera that acted as a major climatic divide, in contrast to the situation on the American continent with the Rocky and Andes Mountains and Eurasia with the Alps and Himalayas. The lack of widespread cordillera means that the climatic pattern of Africa, in many respects, resembles that of the ‘ideal’ or ‘hypothetical’ continent of many textbooks (Goudie, 1996).
A large part of Africa consists of elevated plateaux and these high altitudes play a very important role in decreasing temperatures over extensive areas. The high zonal index of Africa is the key feature resulting in temperature changes. A number of additional factors also play significant roles in temperature changes of the African continent.
6
Firstly, the symmetrical positioning of Africa with regard to latitude, extend almostequidistant from the Equator with its northern extremity at Cape Blanc (37º N) and southern extremity at Cape Agulhas (35º S). Roughly comparable series of climates can be traced northward and southward from the hot, moist Equatorial belt. Sudan has its complement in Zimbabwe, the Sahara in the arid tracts of the Namib and the Kalahari and the Mediterranean coast in the southwest of South Africa. Secondly, Africa is the most tropical of all the continents. Only the extremes reach far enough pole-wards to be directly influenced by the mid-latitude westerly winds and their accompanying instabilities. The Sahara and the belt of country to the south are warmer than any other part of the world of similar size. Thirdly, Africa is strongly influenced by the subtropical anticyclonic belts of both hemispheres and as a result retains widespread areas of dry climate to both the south and north (Goudie, 1996). However, Africa has hemispherical asymmetry in climate, attributable to the fact that Africa north of the Equator is not only broader than Africa to the south of the Equator, but it also lacks a true ocean boundary to the north and northeast, being bordered instead by the great Eurasian continent. Another factor is the small annual variety of temperatures prevalent in Africa. Approximately one-third of Africa experiences an annual range lower than 6 ºC. In the equatorial region there are particular areas where the range is lower than 3 ºC. Mean monthly maximum temperatures are widespread and more than 32 ºC, while approximately 30% of the area is subject to temperatures exceeding 38 ºC. The highest temperatures occur in the Sahara. The average monthly minimum temperatures indicate the effects of latitude, where only relatively small areas in the interior of southern African and the northern Sahara and its margins have values lower than 5 ºC. It is consequently clear that a large part of the continent experiences very high temperatures (Goudie, 1996).
Africa has an annual rainfall of approximately 725 mm, though there is a great range in precipitation with as little as 1 – 2 mm in the driest parts of the Libyan Desert to about 10 000 mm in the Mount Cameroon foothills. Africa exhibits significant deviancies from the standard world pattern of climatic and rainfall distribution (Trewartha, 1981). This can firstly be attributed to the uneven low rainfalls that occur over a large area of eastern tropical-equatorial Africa (especially in Somalia and Kenya) where, on the basis of a windward position and of uplands rising suddenly, one might have anticipated much wetter environments.
7
Secondly is the occurrence of the dry coastal belt in Ghana and Togo on the Gulf ofGuinea in the central part of a coast which otherwise has a tropical wet climate. Thirdly is the area along the Guinea coast with a high-sun (with the sun being directly overhead) secondary rainfall minimum. A fourth anomaly is the restricted latitudinal spread of wet-and-dry climates to the north of the Equator in contrast to the south because of the unusual southward distribution of the Saharan dry climate (Goudie, 1996).
This environmental diversity, the result of Africa’s abundant combinations of climatological, geological and pedological factors, is reflected in the production of Africa’s fauna and flora. These communities have evolved over time as a result of this heterogeneity (Meadows, 1996). The flora itself, occurring as diverse vegetation types (rain forest, savanna, grassland and desert) with distinct distribution patterns, is an indication of enormous, often destructive, climatic changes in the past. This destruction explains Africa’s relative floristic poverty in tropical regions, spectacular disjunction in vegetation distribution, close similarities among species now widely separated geographically and islands of floristic similarity now distantly isolated (Meadows, 1996).
During the late Cretaceous period (135 to 65 million years ago), Africa was still positioned to the south of its present position across the Equator and tropical conditions prevailed mainly in the northern part of the continent. Rain forest dominated the region up to 15º N and 15º S of the Equator, a tropical humid-climate vegetation formation, indicating that the flowering plants had replaced the gymnospermous Glossopteris flora completely (Axelrod and Raven, 1978).
The East African Mountains had still to be formed, so the extension of rain forest from coast to coast is a likely scenario, even though it retained only partial connection with the South American tropical flora after the separation from Africa during the early Cretaceous (Meadows, 1996). Another plant community present on the continent at this time was temperate rain forest, ample indications of which are found in Cretaceous-age fossil-bearing deposits, although, in southern Africa, it appears to have been species-poor and lacking the characteristic Nothofagus (southern beech) element (Van Steenis, 1972), characteristic of the rest of Gondwana at the time.
8
Additional seasonal climates also occurred at this time, which promoted thewidespread occurrence of savanna woodland and sclerophyllous scrub, although severe aridity still had to develop in Africa (Meadows, 1996).
Africa was located approximately at its present latitude by the mid-Tertiary period, about 33.7 million years ago. The major changes this brought on relates to the ‘migration’ of the rain forest flora southwards to inhabit Central Africa, still connecting the west and east coasts (Meadows, 1996). The development of the East African Mountains led to the evolution of montane elements along the mountain spine. For the first time sclerophyll-dominated vegetation is found in the south-western parts of Africa, as verified by the fossil Banke flora at Arnot, in Namaqualand (Scholtz, 1985). Climates that were more seasonal also existed at this time. These supported the general occurrence of savanna woodland and sclerophyllous scrub, even though severe aridity still had not developed in Africa (Meadows, 1996). This is significant, for it documented the arrival of flora that has subsequently diversified to become one of the most species-rich in Africa, if not in the world (Linder et al., 1992).
Aridity was becoming a dominant environmental factor in parts of northern Africa (the proto-Sahara) and in southern Africa by the late Tertiary period (65 – 1.75 million years ago). The arid and semi-arid areas of southern Africa were established because of the isolation of Antarctica (Kennett, 1980), the growth of an Antarctic ice sheet with the resulting global drop in sea-level and the development of the cold Benguela current along the west coast (Siesser, 1980).
The cooler and drier climates reduced the area of rain forest, so much so that savannas became progressively more widely distributed in Africa (Axelrod and Raven, 1978). This change virtually eliminated the temperate rain forest flora from the continent.
Even though Africa is geologically one of the oldest continents and has a fossil biota going back to the origin of life itself, the major modern-day distribution patterns seem largely to have been determined by the dramatic events of the last 65 million years. The contrast between vegetation mapped as major biomes and mean annual precipitation shows prominent similarities and conformities, despite the obvious over-simplification in doing so.
9
The modern-day biomes of Africa therefore have a character which, with theexception of the Equatorial rain forests, is controlled to a greater or lesser degree by seasonal precipitation (Meadows, 1996).
The five major biomes of Africa can be categorized as tropical rain forest, tropical savannas (with or without an overstorey of trees), semi-arid and arid formations, areas of shrubland with a Mediterranean-type climate and the montane regions (Meadows, 1996).
In contrast to these five major biomes, White (1983) divides Africa into eighteen phytochorological units, based mainly on geography. The phytochoria are classified as Regional Centres of Endemism, Archipelago-like Centres of Endemism, Archipelago-like Centres of Extreme Floristic Impoverishment, Transitional Zones and Mosaics and are characterised in terms of the vegetation and rainfall (Table 2.1; Fig. 2.1). White (1979) defines a Regional Centre of Endemism as a phytochorion which has both more than 50% of its species confined to the area and a total of more than 1000 endemic species. The phytochoria regarded as regional centres of endemism fulfil these criteria with only the Sudanian Region, the status of which is still uncertain, as the exception. The regional centres of endemism are separated by transition zones. The Guinea-Congolia/Sudania, Guinea-Congolia/Zambezia and the Kalahari-Highveld Transition Zones are larger than some regional centres of endemism but have few endemic species and the majority of their species also occur in adjacent phytochoria (White, 1983).
(i) Guineo-Congolian Regional Centre of Endemism
Rain forest on well-drained sites and swamp forest on hydromorphic soils initially covered this area. At present little undisturbed rain forest remains while secondary grassland and various stages of forest regrowth are widespread. Small patches of edaphic grassland on certain hydromorphic and other soils not suited to the growth of trees are also present. Stunted forest and various types of bushland and thicket occur in upland areas, above about 1000 m, especially in rocky places. A number of Afromontane species are found in upland areas but it is only on the highest peaks such as Mount Cameroon that they form distinct Afromontane communities which are excluded from the Guineo-Congolian Region.
10
Table 2.1. Classification of the main phytochoria of Africa (White, 1983).
Regional Centres of Endemism Archipelago-like Centre of Endemism Archipelago-like Centre of Extreme Floristic Impoverishment
Regional Transition Zones Mosaics
(i) Guineo-Congolian (viii) Afromontane (ix) Afroalpine (x) Guinea-Congolia/Zambezia (xii) Lake Victoria
(ii) Zambezian (xi) Guinea-Congolia/Sudania (xiii) Zanzibar-Inhambane (iii) Sudanian (xiv) Kalahari-Highveld (xv) Tongaland-Pondoland (iv) Somalia-Masai (xvi) Sahel
(v) Cape (xvii) Sahara
(vi) Karoo-Namib (xviii) Mediterranean/Sahara (vii) Mediterranean
11
Fig. 2.1. White’s (1983) vegetation map of Africa indicating the main phytochoria. (i) Guineo-Congolian Centre, (ii) Zambezian Centre, (iii) Sudanian Centre, (iv) Somalia-Masai Centre, (v) Cape Centre, (vi) Karoo-Namib Centre, (vii) Mediterranean Centre, (viii / ix) Afromontane Archipelago-like Centre / AfroalpineArchipelago-like Region of Extreme Floristic Impoverishment, (x) Guinea-Congolia/Zambezia Transition Zone, (xi) Guinea-Congolia/Sudania Transition Zone, (xii) Lake Victoria Mosaic, (xiii) Zanzibar-Inhambane Mosaic, (xiv) Kalahari-Highveld transition zone, (xv) Tongaland-Pondoland Mosaic, (xvi) Sahel Transition Zone, (xvii) Sahara Transition Zone, (xviii) Mediterranean/Sahara Transition Zone.
12
Most of the Guineo-Congolian Region is relatively dry when compared to rain forestsin other areas and receives between 1600 and 2000 mm of rainfall per year (White, 1983).
(ii) Zambezian Regional Centre of Endemism
The Zambezian Region is Africa’s second largest major phytochorion and shows the widest range of vegetation types and probably the richest and most diversified flora. These vegetation types range from dry forest, swamp and riparian forest, transition woodland, woodland, thicket, scrub woodland to grassland. Almost the whole of the Zambezian Region lies within the tropical summer-rainfall zone, except towards the coast where the climate is continental in character with a noticeably larger seasonal variation in temperature than that of the Guineo-Congolian Region. There is a single rainy season, mainly from November to April, but in some places it may be interrupted by a dry spell lasting for two to three weeks. Rainfall in this region is between 500 and 1400 mm per year, decreasing from north to south, but there are distinct regional variations (White, 1983).
(iii) Sudanian Regional Centre of Endemism
The Sudanian Centre lies inside the northern tropical summer rainfall zone just like the Zambezian Region south of the Equator and the climates of the two are broadly similar especially with regard to rainfall, but temperatures in the Sudanian Region are appreciably higher and because of the harmattan wind, the dry season is more severe than in its southern counterpart. This Centre is bordered in the north by the deserts and semi-deserts of the Saharo-Sindian Region, in to the south it stretches to the tropical forests of the Guineo-Congo Region. The strongly seasonal climate is reflected in the variety of vegetation types from poor thorn scrub and savanna to quite rich deciduous woodland (White, 1983).
(iv) Somalia-Masai Regional Centre of Endemism
The biggest part of this region is covered with deciduous bushland and thicket which grade into and are replaced by semi-evergreen and evergreen bushland and thicket on the lower mountain slopes. There are smaller areas of scrub forest, riparian forest, secondary grassland and wooded grassland, seasonally waterlogged grassland, semi-desert grassland and shrubland and desert.
13
The climate is arid to semi-arid and rainfall is generally less than 500 mm per yearand in some places as low as 20 mm. Throughout the Somalia-Masai Region there are great variations in rainfall from year to year. Most areas have two rainy seasons, separated by periods of drought due to the influence of the SW monsoon in summer and the NE monsoon in winter, but for the most part, these monsoons do not bring rain. Instead, rainfall occurs during the prevailing periods of calm (White, 1983).
(v) Cape Regional Centre of Endemism
Typical Cape vegetation does not inhabit the whole area. Large communities of Karoo and Afromontane vegetation and small patches of bushland of Tongaland-Pondoland affinity are also present. Rainfall ranges from 300 to 2500 mm, but in the mountains it can reach up to 5000 mm per year. The highest winter rainfall is usually during the winter months (White, 1983).
(vi) Karoo-Namib Regional Centre of Endemism
The vegetation of the Namib Desert ranges from Outer Namib fog desert, gravel desert, rocky outcrops, Welwitschia mirabilis transition zone to river-bed communities. The semi-desert vegetation of the Karoo ranges from shrubland, dwarf succulents and succulent shrubs, non-succulent bushes, bushy trees and tall shrubs, grasses, geophytes and annuals to riparian scrub forest. Rainfall in the Namib Desert is less than 100 mm per year and elsewhere it hardly ever exceeds 250 mm. The seasonality of rainfall varies greatly throughout this region and considerable variation in the amount and distribution of rainfall occur from year to year, particularly in the more arid parts. Even in the wetter part of the summer-rainfall belt, winter influences dominates this rainfall pattern approximately one in every twelve years (White, 1983).
(vii) Mediterranean Regional Centre of Endemism
A large part of the Maghreb (northern part of Morocco, Algeria and Tunisia) was formerly covered with forest, but on clay soils in the semi-arid scrub forest, which was then dominated by Olea europaea L. and various types of bushland or thicket. Where the Maghreb was not covered by forest, non-forest woody vegetation occurred and was confined to shallow soils, wind-swept ridges, coastal habitats and the summits of the higher mountains.
14
Precipitation occurs mainly in winter and range between 250 and 1000 mm per year.The summer is hot and dry and is more extreme than that of the Cape Region (White, 1983).
(viii and ix) Afromontane Archipelago-like Regional Centre of Endemism and Afroalpine Archipelago-like Region of Extreme Floristic Impoverishment
The Afroalpine Archipelago-like Region of Extreme Floristic Impoverishment is embedded within the Afromontane Region with a variety of vegetation occurring on the mountains. The vegetation types include forest, bamboo, evergreen bushland and thicket, shrubland, grassland and mixed Afroalpine communities. Very few species of the Afroalpine region are not shared with the Afromontane region. The climate is extremely varied. In the forest belt, mean annual rainfall is typically more than 1000 mm, but is less in drier types of transitional- to lowland vegetation. Above the forest belt precipitation lessens and in the Afroalpine belt of some mountains appears to be much less than 1000 mm per year (White, 1983).
(x) Guinea-Congolia/Zambezia Regional Transition Zone
The largest part of this Transition Zone is occupied at present by secondary grassland and wooded grassland dominated almost exclusively by Zambezian species. Other vegetation types include drier peripheral semi-evergreen rain forest, dry evergreen forest, transition woodland and coastal mosaic. In most of this region the climate is similar to that of the Guineo-Congolian and Zambezia Regions.
The dry season is more severe than in the Guineo-Congolian Region but less so than in the Zambezian Region. Rainfall declines rapidly near the Atlantic coast to below 800 mm per year, but relative humidity during the dry-season is high (White, 1983).
(xi) Guinea-Congolia/Sudania Regional Transition Zone
The larger part of this Transition Zone is covered with secondary grassland and secondary wooded grassland. Various types of forest were previously widespread in this area, but have been extensively destroyed by fire and cultivation, peripheral semi-evergreen rain forest being the only survivor. Almost everywhere the climate is intermediate between those of the Guineo-Congolian and Sudanian Regions.
15
A narrow strip of coastal plain in West Africa, extending from Ghana eastward to theBenin Republic, has an anomalously dry climate. In the driest part, near Accra, rainfall is only 733 mm per year (White, 1983).
(xii) Lake Victoria Regional Mosaic
This Regional Mosaic is the meeting-place of five distinct phytochoria, namely Guineo-Congolian, Sudanian, Zambezian, Somalia-Masai and Afromontane Regions. The vegetation comprises a mosaic of floristically poor variants of the characteristic vegetation of the first four regions which are semi-evergreen Guineo-Congolian rain forest, transitional rain forest, swamp forest and scrub forest and in some cases with an admixture of Afromontane species. Climatic gradients are frequently steep and are related to the complex physiography and the distance from Lake Victoria, which is a significant source of precipitation. Around the lake the rainfall is sufficiently high, 1500 to 2000 mm per year, and well distributed throughout the year to support the rain forest. More distanced form the lake, the rainfall is too low to support rain forest and not adequately seasonal to support woodland (White, 1983).
(xiii) Zanzibar-Inhambane Regional Mosaic
Forest once was the most widespread vegetation, but has largely been replaced by secondary wooded grassland and cultivation. There are also extensive areas of scrub forest, edaphic grassland and smaller areas of transition woodland, bushland and thicket. Rainfall is typically between 800 and 1200 mm per year and there is a well-defined dry season. Appreciably higher rainfall is experienced in a few places such as the East Usambara Mountains at Amani with 1946 mm and on the islands of Zanzibar and Pemba where 1964 mm falls at Wete. In these places the amount and distribution of precipitation are enough to support rain forest (White, 1983).
(xiv) Kalahari-Highveld Regional Transition Zone
This zone separates the Zambezian and Karoo-Namib Regional Centres of Endemism and occurs mainly on the great Interior Plateau of southern Africa. Vegetation is complex and can range from scrub forest, riparian scrub, rupicolous bushland and shrubland, wooded grassland to grassland.
16
Rainfall is in-between that of the Zambezian and Karoo-Namib Regions, rangingbetween 250 and 500 mm per year, increasing somewhat in the east towards the Drakensberg. Most of the rainfall occurs in summer (White, 1983).
(xv) Tongaland-Pondoland Regional Mosaic
Vegetation consists of a complex mosaic of forest, scrub forest, evergreen and semi-evergreen bushland and thicket in a matrix of secondary grassland and wooded grassland where it has not been completely destroyed. There are small patches of woodland in the north and edaphic grassland and swamp forest on the coastal plain. Due to the ameliorating effect of the warm Mozambique Current the coastal regions have a moderately high and well-distributed rainfall (White, 1983).
(xvi) Sahel Regional Transition Zone
The Sahel Zone occupies a band which stretches from the Atlantic Ocean on the east to the Red Sea on the west with the Saharan Desert that borders on the north and the Central African rainforests on the south. This band forms a flat or gently undulating landscape below 600 m. Wooded grassland in the south and semi-desert grassland in the north of the Sahel are supported by the extensive sand sheet. Bushland is much more restricted and is mostly confined to rocky outcrops. Numerous types of scrub forest and bushland also occur. Rainfall is unreliable and varies between 150 and 500 mm per year but rises to more than 1000 mm on Jebel Marra. Most rain falls in the three to four summer months, while the dry season is long and severe (White, 1983).
(xvii) Sahara Regional Transition Zone
The vegetation of this Transition Zone ranges from dwarf shrubland, grassland to absolute desert. Three climate zones, northern, central and southern, can be recognized on the basis of rainfall distribution. In the northern zone the rain falls during the cold season with two maxima, in autumn and spring. Even though rain falls every year, there is considerable variation from one year to the next both in distribution and amount. Rainfall declines rapidly toward the south. Rain is episodic in the central region, with the mean annual rainfall being less than 20 mm, except in the high mountains. The southern part receives more rain than the latter two regions and is subject to summer rain (White, 1983).
17
(xviii) Mediterranean/Sahara Regional Transition ZoneThere is considerable change in the vegetation from west to east. In western Morocco the predominant types of vegetation are scrub forest and bushland. A mosaic of grassland dominates the landscape from eastern Morocco to Tunisia. Rainfall is between 100 and 250 mm per year. Precipitation is typically concentrated in the winter months, but in the rain shadow of the Atlas Mountains and on the High Plateau the main peaks are in spring and autumn, but the rainfall may be uneven through the year (White, 1983).
It is hypothesised that climatic stability is an important determinant of a high degree of endemism and that the degree of endemism is not coupled with high rainfall. It is also stated that areas that are consistently arid are just as likely to have endemics as areas that are consistently mesic (Linder, 2001).
When low levels of orbitally forced range dynamics (ORD) (climatic shifts leading to large changes in a species geographical distribution) occur it enables the survival of palaeoendemics as well as the preservation of genetic variation within those persisting populations, whereas high levels of ORD reduces endemism (Jansson, 2003). It is also said that the smaller the climatic shifts in an area, the more likely that neoendemic species will evolve and that it is more probable that new clades will persist and not go extinct (Jansson, 2003).
2.2. The family Apocynaceae
The Apocynaceae was first described by Jussieu in 1789 under the name Apocineae. Robert Brown (1810) separated the Asclepiadeae from Apocineae in his publication “On the Asclepiadeae”. The main reason for this separation was the absence of translators (complex pollen carriers formed from hardened secretions of the style-head) and complex pollinia in the Apocineae and the presence thereof in the Asclepiadeae. This proved to be an incongruous character as demonstrated by Endress and Bruyns (2000) when they found that the presence and/or absence of pollinia and translators form a continuum between the two extremes in the modern Apocynaceae (Apocineae).
18
The relationship between the Apocynaceae and Asclepiadaceae, as defined byBrown (1810) has always been the subject of taxonomical controversy due to their obviously close relationship. Attempts to resolve this problem included combining them as an order separate from the Gentianales, as one family (Judd et al., 1994), or as a suborder within the Gentianales (Endress and Bruyns, 2000). After careful consideration of the distribution of character states of various morphological characteristics between the two families, Endress (1997) concluded that the delimitation between the two families is artificial, because there is a gradation of many character states from one family to the other. Endress and Bruyns (2000) therefore proposed that these two families be united.
Central in modern classification is the establishment of monophyletic taxa. Molecular analyses have indicated that the Apocynaceae sensu stricto (as in Brown, 1810) is paraphyletic when excluding the Asclepiadaceae (Judd et al., 1994; Civeyrel, 1996; Endress et al., 1996; Sennblad and Bremer, 1996; Sennblad et al., 1998; Endress and Bruyns, 2000; Sennblad and Bremer, 2000; Endress, 2001; Endress and Stevens, 2001). Combining the Apocynaceae sensu stricto and Asclepiadaceae into one family is the most effective way to achieve a monophyletic family. The cladistical results of Livshultz et al. (2007) also support this conclusion.
At present, the Apocynaceae is circumscribed in the sense of Jussieu, therefore including Brown’s (1810) Asclepiadaceae and is referred to as the Apocynaceae sensu lato (Endress, 2004). For the sake of clarity, Brown’s (1810) Apocynaceae is called the Apocynaceae sensu stricto.
Robert Brown was unquestionably one of the most significant people regarding the early taxonomy of the Apocynaceae. He described over 40 of the 53 genera comprising the Apocynaceae and Asclepiadaceae of the time, the majority of which are still valid today (Meve, 2002). Thus far, the estimated number of genera in this family has grown to 357 with roughly 5100 species (Nazar et al., 2013), more than seven times the number of genera and 30 times the number of species that was known to Robert Brown.
19
2.3. The Subfamily Periplocoideae2.3.1. Circumscription of the Periplocoideae
Brown (1810) proceeded to subdivide his new family, Asclepiadeae (presently the Asclepiadaceae), into three groups or “tribes”, namely Periploceae, Asclepiadeae verae and an unnamed group containing only the genus Secamone R.Br., based on morphological differences and similarities, primarily the number of pollinia per flower and the type of translator. The ‘Periploceae’ was characterised by pollen clustered in tetrads (rarely in pollinia), each anther containing four pollen sacs and the tetrads (or two pollinia) from the thecae of two adjacent anthers deposited onto a spoon-like translator.
The unnamed group containing Secamone differed in that pollen occur in pollinia, each anther is divided into four pollen sacs, producing four pollinia, and two pollinia from each of the thecae of two adjacent anthers are attached onto a clamp-like translator. The ‘Asclepiadeae verae’ also has pollen in pollinia, but each anther has only two pollen sacs producing two pollinia. One pollinium from each of the thecae of two adjacent anthers is attached onto a clamp-like translator.
Schlechter (1905) elevated Brown’s Periploceae to family level with the name Periplocaceae. Botanists such as Bullock (1956), Kunze (1990; 1993), Dave and Kuriachen (1991), Liede and Kunze (1993), Nilsson et al. (1993) and Swarupanandan et al. (1996), concurred with the idea that the Periplocaceae is a separate family, closely related to the Asclepiadaceae.
However, after extensive morphological research, Endress and Bruyns (2000) concluded that there is little support for separating the Periplocaceae from the remainder of the ascleps at family level. They proposed that the Periplocaceae be reduced to subfamily status within the Asclepiadaceae. Molecular studies using the rbcL gene region did not support the monophyly of the Asclepiadaceae and Periplocaceae as separate families but rather indicated a close relationship between the Apocynaceae sensu stricto and Periplocaceae (Kunze, 1990; 1996; Judd et al., 1994; Struwe et al., 1994; Sennblad and Bremer, 1996; Endress, 1997; Sennblad, 1997).
20
The bulk of the new information on relationships among genera is based on DNAdata, especially since it has continually revealed convergences that were not realised as such in earlier classifications based on morphological data alone (Endress, 2004).
Various phylogenetic analyses did support the grouping of the genera of the Apocynaceae sensu lato into five subfamilies in accordance with Endress and Bruyns (2000), as well as the monophyly of the Periplocoideae (Sennblad and Bremer, 1996; Civeyrel et al., 1998; Potgieter and Albert, 2001; Ionta and Judd, 2007; Lahaye et al., 2007; Ionta, 2009). However, the studies of Sennblad and Bremer (1996), Potgieter and Albert (2001) and Livshultz et al., 2007 also show that the Periplocoideae are likely nested within the paraphyletic Apocynoideae clade. Fishbein’s study (2001), based on the chloroplast matK gene, confirms the monophyly of the subfamilies where he assumes a closer relationship of Periplocoideae to non-asclepiad Apocynaceae than to the remainder of the Asclepiadaceae.
Today, two centuries later, Brown’s original three groups within the Asclepiadaceae are thus reinstated and defined using his morphological criteria, supported by molecular data as representing monophyletical groups or subfamilies, these being the Periplocoideae, Secamonoideae and Asclepiadoideae (Endress, 2001; 2004). Periploca (Linnaeus, 1753) was the first genus described in the Periplocoideae. Brown (1810; 1819) then added Cryptolepis, Gymnanthera, Hemidesmus and Cryptostegia, Blume (1825 – 1826; 1828) added Leposma, Phyllanthera and Lepistoma, Wallich (1832) contributed Finlaysonia, while Wight and Arnott (Wight, 1834) added Decalepis and Streptocaulon. With the exception of Periploca, all these genera were described from specimens collected in India and Indonesia. The first genera described from Africa were Ectadium by E. Meyer (1837), Raphionacme by Harvey (1842) and Aechmolepis, Tacazzea and Suchellia by Decaisne (1844). From Madagascar, the third significant centre of diversity for the Periplocoideae, Decaisne (1844) described Camptocarpus, Harpanema and Pentopetia.
21
To date, 86 generic names by 30 authors have been contributed to thePeriplocoideae. The number of recognized genera in this subfamily has changed considerably throughout the years and is still in flux. Venter and Verhoeven (1997) proposed a tribal classification of the Periplocoideae that included 44 genera. Since then two of the smaller genera were placed in synonymy by Klackenberg (1997) and Venter and Verhoeven (2001), reducing the number of genera to 42. The number of species is now approximately 190, so that the number of genera, approximately 42, is proportionally high. Given the small size of the subfamily, an excessively high number of monotypic and ditypic genera have been described. About 117 species belongs to six major genera, while most of the other genera are uni- or bispecific. The largest genera are Raphionacme (36 species and 2 subspecies), Cryptolepis R.Br. (30 species), Pentopetia Decne. (23 species), Periploca (13 species), Camptocarpus Decne. (9 species) and Streptocaulon Wight & Arn. (9 species) (Venter, 2009).
These changes may be attributable to a variety of factors such as the significant morphological diversity of the group, the undercollection of numerous taxa (particularly in Asia) which are therefore poorly known (Venter and Verhoeven, 1997; Meve and Liede, 2004) and the small yet highly complex flowers that are difficult to interpret from herbarium material.
2.3.2. Distribution and habitat of the Periplocoideae
When looking at the distribution of the Periplocoideae, this subfamily is limited to the Old World, which is in Africa, Madagascar, Europe, Asia and Australia, (Venter et al., 1990a; Venter, 1997; Venter and Verhoeven, 1997). Most of the genera and species are woody climbers, some very large (Mondia Skeels, Myriopteron Griff. and Tacazzea), inhabiting tropical rain and swamp forests or, rarely, temperate forests. A small number of species, such as Raphionacme galpinii and R. hirsuta, inhabit the grasslands of Africa and a few are erect or straggling shrubs and are found in arid semi-desert and desert habitats (e.g., the three Ectadium species, Periploca aphylla Decne., P. visciformis (Vatke) K.Schum., Raphionacme haeneliae and R. namibiana (Venter and Verhoeven, 1986; 1996; Venter et al., 1990b; Venter, 1997).
22
Interestingly, no members of the Periplocoideae inhabit the South African CapeFloral Kingdom with its winter rainfall. Nevertheless, a few species, like Periploca angustifolia Labill. and P. gracilis Boiss., occur in the Mediterranean macchia of North Africa and Europe (Venter, 1997).
Periplocoideae are never dominant in the plant communities where they occur, although lianas like Mondia and Tacazzea may be prominent because of their size. The taxa are often widely distributed, but large numbers of individual plants seldomly occur (Venter and Verhoeven, 1988).
The exception is Cryptostegia grandiflora Roxb. ex R.Br. of Madagascar, which has become common in disturbed locations in Madagascar. In Queensland, Australia, it has become a major noxious weed of grazing lands (Marohasy and Forster, 1991).
2.4. The genus Raphionacme
2.4.1. Historical notes on Raphionacme
The name Raphionacme was established by William Henry Harvey in 1842 when he described a number of new genera from South Africa. He gave no indication as to the origin of the name, but Raphionacme could be an inflection of Raphanus (radish) due to the similarity between the tubers of the two genera (Venter, 2009).
Variations of this generic name have been proposed, for instance Rhaphiacme or Raphiacme (Schumann, 1893; 1895) and Rhaphidonacme or Raphidacme, but the International Code of Botanical Nomenclature (ICBN) (Greuter et al., 2000) confirmed the name Raphionacme.
Harvey described two species, R. zeyheri and R. divaricata, in his new genus Raphionacme. In his revisions of Raphionacme, N.E. Brown (1902; 1907) regarded an older species, Brachystelma hirsutum E.Mey., as synonymous with R. divaricata Harv., a dubious conclusion (Venter, 2009). However, Dyer (1942) accepted Brown’s synonymy and B. hirsutum E.Mey. being the basionym, he changed the species name to R. hirsuta. Phillips (1951) then designated R. hirsuta as the generic type.
23
Raphionacme is the largest genus in the Periplocoideae. Various authors havedescribed new Raphionacme species with the main contributions by N.E. Brown,
F.F.R. Schlechter, H.J.T. Venter and R.L. Verhoeven and to a lesser extent S. Moore, K. Schumann and A.A. Bullock, resulting in about 70 published species
names. According to the latest revision by Venter (2009) the genus consists of 36 species and 2 subspecies, with the other names placed in synonymy.
2.4.2. Sectional classification of Raphionacme
Raphionacme was divided into four sections by Schumann (1895) based on the ten species known to him. When Venter and Verhoeven (1988) classified the 31 species of Raphionacme known at the time, they used the names given by Schumann (1895) and amended the descriptions of the sections (Table 2.2). These sections were based on morphological structure of inflorescences, flowers and growth habit.
2.4.3. Distribution and habitat of Raphionacme
Raphionacme is virtually endemic to Africa, with only one species, R. arabica, occurring on the Arabian Peninsula. Raphionacme is absent from the Sahara Desert, the winter rainfall region of the western/southwestern Cape Region of South Africa, the very wet tropical forests of western Africa and virtually excluded from the Mediterranean winter rainfall region of North Africa (except two collections of R. splendens). The highest concentration of Raphionacme species is found between 15 – 20º S and 25 – 35º E, comprising northern Botswana, Zimbabwe, southwestern Zambia and northwestern Mozambique (Venter, 2009).
In his biogeographical analysis of the flora of Oman, Ghazanfar (1992) uses White’s (1983) classification and Leonard’s (1988) extension of the phytochoria in Asia. The Dhofar Mountains in the southeast of Yemen form a steep, arid escarpment rising above the coastal plain bordering the Arabian Sea.
24
Table 2.2. Raphionacme species within Venter and Verhoeven’s (1988) foursections. Raphionacme sect. Raphionacme Raphionacme sect. Cephalacme K.Schum. Raphionacme sect. Speiracme K.Schum. Raphionacme sect. Pseudochironia K.Schum. R. dyeri R. angolensis R. flanaganii R. bingeri
R. hirsuta R. galpinii R. keayii R. brownii R. lanceolata R. globosa R. longifolia R. caerulea
R. longituba R. utilis R. monteiroae R. chimanimaniana R. madiensis R. vignei R. procumbens R. excisa
R. palustris R. sylvicola R. grandiflora R. velutina R. welwitschii R. linearis
R. zeyheri R. michelii
R. namibiana R. splendens R. sp. nov. = R. borenensis
25
The climate and floristic composition correspond to the Somalia-Masai RegionalCentre of Endemism of Africa (White, 1983) and is also influenced by the southwestern Monsoon (Ghazanfar, 1992). It would seem as if Raphionacme arabica forms part of the concentration of endemic species found in the escarpment woodlands of Dhofar.
Raphionacme plants are usually found in savanna and grassland where the climate ranges from arid to seasonally wet (Venter and Verhoeven, 1988; Venter, 2009). A few exceptional species can be found in seasonally dry swamps (R. linearis, R. palustris and R. splendens) (Venter, 2009) or moist and dense escarpment woodland but with a marked dry season (R. sylvicola) (Pers. com. Dr David Goyder, Herbarium Royal Botanic Gardens, Kew, Richmond, United Kingdom). The members of the genus are always lesser components of the communities in which they occur (Venter and Verhoeven, 1988; Venter, 2009).
2.4.4. Morphology of Raphionacme
The genus consists of latex bearing geophytes, with most species being prostrate or erect herbs and a small number being climbers (Venter, 2009). A single subterranean taproot tuber is usually turnip-shaped, with one to a few subterranean, erect, perennial stems occurring at the crown of the tuber.
Annual (rarely perennial) aerial stems develop from these subterranean stems and can be twining, erect or procumbent, with dichotomous or lateral branching. Interpetiolar ridges with reddish turbinate colleters occur on the stems.
Opposite leaves are sessile to petiolate, with blades broadly to narrowly ovate, elliptic, obovate or linear. Leaf texture varies from herbaceous to leathery, the surface can be glabrous or hairy, the main veins are prominently visible abaxially and the secondary veins are arching, divaricate or patent.
The cymose inflorescences occur terminally or pseudo-axillary and can be raceme-like, plume-raceme-like, corymbose-like or globose in appearance, either few or many flowered, lax or compact.
26
The species of Raphionacme are distinguishable by their floral morphology. Flowersare typically pentamerous and actinomorphic, bisexual and semi-epigynous. The free sepals may or may not have paired colleters at inner bases with the sepal shapes ranging from ovate to triangular. In most species the corolla is glabrous inside but can be either glabrous or hairy outside. The corolla is divided into a distinct upper tube (that is the portion above the circle of nectaries) and a very short lower tube that is annular around the upper half of the ovaries. The shape of the upper tube varies from campanulate to cylindrical with the inner face vertically fluted due to coronal ridges. The corona lobes, which can be spreading to reflexed, are ovate, obovate or triangular. The corona arises from the corolla mouth with the lobes opposite the sepals. The corona consists of five lobes, borne in the corolla mouth, attached to coronal feet, fused to the corolla just below the corolla lobe sinuses. The simple or variously incised corona can be glabrous or hairy. Five stamens, that dehisce laterotrorsely with full or terminal half-length slits, arise from the coronal feet and are glabrous with free filaments. A gynostegium is formed by ovate, oblong-ovate, triangular or hastate shaped anthers which are fused to the style-head. The gynostegium is found in the corolla mouth or could be elevated above it. Pollen is usually in tetrads with single grains being 8 – 16-porate. Pocket-like, deep green nectaries are found at the base of the upper corolla tube and are fused to the coronal feet via coronal ridges. The two semi-inferior ovaries are unilocular and many ovuled. The two styles are basally free, but fuse into one compound style that is terete or rarely fluted, glabrous or hairy and becomes terminally enlarged to form a pentagular style-head which is broadly ovoid, oblong-ovoid or deltoid.
Five translators, each consisting of a receptacle (spoon-like structure) with a stalk and viscous disc, are embedded into the upper surface of the style-head. The follicles are erect or divaricate, narrowly ovoid to cylindrical and can be single or in pairs. A coma or a ring of hairs occur on the obliquely ovate seeds. (Venter, 2009)
27
2.5. Molecular analysis2.5.1. Ribosomal DNA (rDNA)
Multiple copies of ribosomes are important for the functioning of organisms. The ribosome genes exist in tandem arrays composed of thousands of copies per array. The ribosomes translate mRNA to build polypeptide chains, thus, making the ribosomes important structures in a cell (Baldwin et al., 1995; Wendel et al., 1995; Cronn et al., 1996; Soltis et al., 1997; Kuzoff et al., 1998; Soltis and Soltis, 1998; Small et al., 2004; Poczai and Hyvönen, 2010).
The rDNA locus have virtually the same structure within a wide variety of taxa, but due to functional constraints, it is commonly assumed that all ribosomal copies present in the genome have fairly identical sequences (Small et al., 2004). Coding regions, like the small and large subunit genes of the ribosomes, signify some of the most conserved sequences in eukaryotes as a result of a strong selection for the regions to prevent against any loss of function (Hillis and Dixon, 1991; Poczai and Hyvönen, 2010).
Since the specific regions of the rDNA loci are conserved differently, the sequence information from the different rDNA locus can be utilized at different taxonomic levels. The internal spacer regions can be beneficial from generic to species level and are now commonly used in phylogenetic studies (Poczai and Hyvönen, 2010).
2.5.2. Characterising the Internal Transcribed Spacer (ITS)
The ITS region is located in the 18S-5.8S-26S region and consists of three components: two spacer regions (ITS-1 and ITS-2) and a highly conserved 5.8S rDNA exon (Fig. 2.2) (Baldwin et al., 1995; Li et al., 2011). The spacer sequences of ITS-1 and ITS-2 are both incorporated into the mature ribosome. The ribosomal RNA undergoes a specific splicing process during maturation, before mRNA translation (Poczai and Hyvönen, 2010).
The 5.8S subunit is almost constant in length (mostly 163 or 164 bp) in reported
angiosperms. The whole ITS region in angiosperms seems to be between 500 – 750 bp in length. In other seed plants, it can be up to 1500 – 3500 bp.
28
Fig. 2.2. Schematic representation of the universal structure of the rDNA region inplants as adopted from Poczai and Hyvönen (2010). The small subunit (18S) and large subunit genes (5.8S and 26S) are separated by the internal transcribed spacer 1 (ITS-1) and internal transcribed spacer 2 (ITS-2).
